Shopify–ChatGPT integration fuels 257% jump in merchant-agent revenue

Conversational commerce is finally getting a real scorecard. Not a demo, not a concept slide — actual sessions, orders, and revenue. The new wrinkle is Shopify’s connection into ChatGPT, which lets merchants surface live product data inside AI shopping flows and send buyers through Shopify checkout. Now a merchant tool vendor, Tuki, says one Shopify brand saw agent-attributed revenue jump 257% after the rollout. (shopify.com) ### What actually changed? The big shift is that ChatGPT is no longer just helping people research products. OpenAI has been building shopping and checkout features directly into ChatGPT, while Shopify has been wiring merchant catalogs, pricing, inventory, images, and variants into those experiences. That means a shopper can discover a product in chat, get current product info, (shopify.com)o messy. (openai.com) ### Why is Shopify the key piece? Because Shopify already sits underneath a huge chunk of online retail plumbing. Its pitch is basically: set up product data once, then syndicate it across AI channels. Shopify now says Agentic Storefronts can push merchant products into ChatGPT and other AI surfaces from the Shopify admin, with checkout still running through Shopify. That matters because (openai.com) assistant. (shopify.com) ### What is the 257% number? The number comes from a 90-day comparison highlighted by Tuki and echoed in broader industry coverage. The example: 2,794 agent-referred sessions, 150 attributed orders, $15,138 in revenue, and a 5.9% conversion rate for a high-end fashion brand — up 257% versus the previous 90 days. That is still one merchant example, not a market-wide benchmark, but i(shopify.com)commerce reporting instead of AI hype. (findskill.ai) ### Is that a big number? Yes — but in a very specific way. $15,138 is not huge in absolute ecommerce terms. The interesting part is the efficiency. A 5.9% conversion rate from agent-referred sessions is strong for a new channel, especially one that is still early, limited, and not yet a default shopping habit for most consumers. Think of it less like a giant ne(findskill.ai)arriving closer to purchase. (findskill.ai) ### Why would chat convert better? Because the assistant can collapse the messy middle. Normal shopping means search results, tabs, filters, reviews, and abandoned carts. In chat, the buyer can ask for “black running shorts under $60 that don’t ride up,” get narrowed options, compare them side by side, and move straight to purchase. OpenAI’s shopping product is (findskill.ai) data gives the model cleaner inventory to work with. (openai.com) ### What’s the catch? Attribution is still squishy. “Agent-attributed” does not automatically mean the AI channel created all the demand from scratch — some of that demand may have been captured, assisted, or accelerated. And one merchant case does not prove the same result for beauty, electronics, supplements, or low-consideration goods. The infrastructure is real. The performance evidence is early. (findskill.ai) ### Why does this matter now? Because the fight is shifting from “can AI recommend products?” to “who owns the transaction layer?” Shopify is trying to be the commerce backbone for AI channels, not just for storefront websites. OpenAI is turning ChatGPT into a place where shopping can start and finish. If merchant economics hold up, acquisition paths, SEO strategy, and marketplace power all start to move. (shopify.com) ### Bottom line? The 257% jump is not proof that AI shopping has already won. But it is proof of something more useful — the stack is live enough to produce merchant revenue that can be measured. Once that happens, this stops being a product announcement and starts being a channel test. (findskill.ai)